Related thread where the article runs chatGPT through a number of political alignment tests. (Didn't get as much attention though).

https://news.ycombinator.com/item?id=34038485

That's a more in-depth article, so I'm going to clone that thread (it's too old to re-up directly) and move the comments thither.

Edit: comments moved to https://news.ycombinator.com/item?id=34181339.

Looks simply perfect.

~30% left and ~30% libertarian sounds like reasonably caring, reasonably pro-freedom, but not too fanatic like in "communism&chaos or death".

Is it fair to say this more indicative of Open AI’s political leanings? Based on selected training sets etc.